Despite significant improvements in intensive care and dialytic technology over the past two decades, outcomes associated with acute renal failure (ARF) have remained poor. There is wide regional and intra-individual variation in the early management of ARF (monitoring, fluid resuscitation, diuretic and pressor use), and in the provision of dialysis (timing, modality, frequency, dose and duration). Furthermore, relatively little attention has been paid to the nature of renal injury in ARF, and to relevant outcomes other than survival (recovery of renal function, functional status and quality of life, length of hospital stay, costs). This project is part of an Interactive Research Project Grant (IRPG) application which will attempt to identify key modifiable elements in the evaluation and management of ARF. The specific aims of this application are (a) to refine and validate methods assessing the severity of renal dysfunction, (b) to develop disease-specific severity of illness and co-morbidity instruments valid in patient with ARF, (c) to link elements in the process of care (e.g. timing and modality of dialysis) to outcomes, adjusting for confounding and bias using established and novel statistical methods, and (d) to evaluate the feasibility of influencing the pre-dialytic management of ARF patients at these and other IRPG sites. Data will be shared among all IRPG sites to enhance the statistical power and generalizability of observational analyses. The severity of illness and co-morbidity indices will be used in an effort to improve prognostic accuracy, to define criteria for withholding and withdrawing dialysis because of recovery of renal function or futility, and ultimately, to provide risk adjustment in the evaluation of alternative treatment strategies aimed at improving survival, recovery of renal function and other outcomes in ARF. It is suggested that recognition of the optimal predialytic management strategies in ARF will reduce variation and improve the overall quality of care. It is anticipated (by the applicants) that information gained from this study, being largely observational, will be used to rationally design intervention strategies in future years.